Pub Date : 2021-12-17DOI: 10.1109/ICECE54449.2021.9674233
Yang Liu, Jiaze Zhang, Shengmao Zhang, Fei Wang, Xueseng Cui, Zuli Wu, Guohua Zou, Jing Bo
The fish target detection algorithm lacks a good quality data set, and the algorithm achieves real-time detection with lower power consumption on embedded devices, and it is difficult to balance the calculation speed and identification ability. To this end, this paper collected and annotated a data set of 84 fishes containing 10042 images, and based on this data set, proposed a multi-scale input fast fish target detection network (BTP-yoloV3) and its optimization method. The experiment uses Depthwise convolution to redesign the backbone of the yoloV4 network, which reduces the amount of calculation by 94.1%, and the test accuracy is 92.34%. Then, the training model is enhanced with MixUp, CutMix, and mosaic to increase the test accuracy by 1.27%; Finally, use the mish, swish, and ELU activation functions to increase the test accuracy by 0.76%. As a result, the accuracy of testing the network with 2000 fish images reached 94.37%, and the computational complexity of the network BFLOPS was only 5.47. Comparing the YoloV3∼4, MobileNetV2- yoloV3, and YoloV3-tiny networks of migration learning on this data set. The results show that BTP-Yolov3 has smaller model parameters, faster calculation speed, and lower energy consumption during operation while ensuring the calculation accuracy. It provides a certain reference value for the practical application of neural network.
{"title":"Research on Optimization Method of Multi-scale Marine Fish Target Fast Detection Network","authors":"Yang Liu, Jiaze Zhang, Shengmao Zhang, Fei Wang, Xueseng Cui, Zuli Wu, Guohua Zou, Jing Bo","doi":"10.1109/ICECE54449.2021.9674233","DOIUrl":"https://doi.org/10.1109/ICECE54449.2021.9674233","url":null,"abstract":"The fish target detection algorithm lacks a good quality data set, and the algorithm achieves real-time detection with lower power consumption on embedded devices, and it is difficult to balance the calculation speed and identification ability. To this end, this paper collected and annotated a data set of 84 fishes containing 10042 images, and based on this data set, proposed a multi-scale input fast fish target detection network (BTP-yoloV3) and its optimization method. The experiment uses Depthwise convolution to redesign the backbone of the yoloV4 network, which reduces the amount of calculation by 94.1%, and the test accuracy is 92.34%. Then, the training model is enhanced with MixUp, CutMix, and mosaic to increase the test accuracy by 1.27%; Finally, use the mish, swish, and ELU activation functions to increase the test accuracy by 0.76%. As a result, the accuracy of testing the network with 2000 fish images reached 94.37%, and the computational complexity of the network BFLOPS was only 5.47. Comparing the YoloV3∼4, MobileNetV2- yoloV3, and YoloV3-tiny networks of migration learning on this data set. The results show that BTP-Yolov3 has smaller model parameters, faster calculation speed, and lower energy consumption during operation while ensuring the calculation accuracy. It provides a certain reference value for the practical application of neural network.","PeriodicalId":166178,"journal":{"name":"2021 IEEE 4th International Conference on Electronics and Communication Engineering (ICECE)","volume":"158 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121682867","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-12-17DOI: 10.1109/ICECE54449.2021.9674567
Jovial Niyogisubizo, L. Liao, Yuyuan Lin, Linsen Luo, Eric Nziyumva, Evariste Murwanashyaka
Crash injury severity prediction is a promising area of interest in traffic safety and management. Recently, machine learning approaches are becoming popular due to their ability to enhance the prediction performance through the bias-variance trade-off-technique. However, some of these methods are criticized to perform like a ‘black box’ approach while predicting and analyzing crash injury severity and produce low accuracy. In this study, we propose a novel stacking framework based on a hybrid of Gradient Boosting (GB), Adaptive Boosting (AdaBoost), and Multilayer Perceptron (MLP) to predict accurately crash injury severity. On the traffic collision dataset provided by the Seattle City Department of Transportation from 2004 to 2021, the proposed model has demonstrated superior performance when compared with the base models. Furthermore, SHAP (SHapley Additive exPlanation) is used to interpret the contribution of every feature on model performance and provide recommendations to responsible authorities.
{"title":"A Novel Stacking Framework Based On Hybrid of Gradient Boosting-Adaptive Boosting-Multilayer Perceptron for Crash Injury Severity Prediction and Analysis","authors":"Jovial Niyogisubizo, L. Liao, Yuyuan Lin, Linsen Luo, Eric Nziyumva, Evariste Murwanashyaka","doi":"10.1109/ICECE54449.2021.9674567","DOIUrl":"https://doi.org/10.1109/ICECE54449.2021.9674567","url":null,"abstract":"Crash injury severity prediction is a promising area of interest in traffic safety and management. Recently, machine learning approaches are becoming popular due to their ability to enhance the prediction performance through the bias-variance trade-off-technique. However, some of these methods are criticized to perform like a ‘black box’ approach while predicting and analyzing crash injury severity and produce low accuracy. In this study, we propose a novel stacking framework based on a hybrid of Gradient Boosting (GB), Adaptive Boosting (AdaBoost), and Multilayer Perceptron (MLP) to predict accurately crash injury severity. On the traffic collision dataset provided by the Seattle City Department of Transportation from 2004 to 2021, the proposed model has demonstrated superior performance when compared with the base models. Furthermore, SHAP (SHapley Additive exPlanation) is used to interpret the contribution of every feature on model performance and provide recommendations to responsible authorities.","PeriodicalId":166178,"journal":{"name":"2021 IEEE 4th International Conference on Electronics and Communication Engineering (ICECE)","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114754848","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-12-17DOI: 10.1109/ICECE54449.2021.9674505
Yitong Xiong, Xiaoping Zeng, Y. Pu, Enze Nie
This paper presents an mm-wave CMOS active phase shifter with a low-loss I/Q generator for 5G applications. A hybrid high-pass/low-pass network topology is proposed to build the I/Q generator. The I/Q generator exhibits lower insertion loss compared to traditional ones. The designed phase shifter is constituted by an active single-to-differential convertor, a fully differential vector-summing core and a differential-to-single convertor. The mm-wave phase shifter, fabricated in 90 nm CMOS technology, provides 360° of phase shifting range over 35-43 GHz while consuming only 17.3 mW of power and 0.36 mm2 of chip area (excluding pads). Measurement results show that the phase shifter achieves an average gain of1.6-5 dB over 35-43 GHz. RMS phase error and RMS amplitude error for 32 measured phase states are less than 1 dB and 5.1°. The proposed phase shifter, which is able to cover the released frequency bands of n260, is a good candidate for 5G mm-wave applications.
{"title":"An mm-wave CMOS Active Phase Shifter with a Low-Loss I/Q Generator for 5G Applications","authors":"Yitong Xiong, Xiaoping Zeng, Y. Pu, Enze Nie","doi":"10.1109/ICECE54449.2021.9674505","DOIUrl":"https://doi.org/10.1109/ICECE54449.2021.9674505","url":null,"abstract":"This paper presents an mm-wave CMOS active phase shifter with a low-loss I/Q generator for 5G applications. A hybrid high-pass/low-pass network topology is proposed to build the I/Q generator. The I/Q generator exhibits lower insertion loss compared to traditional ones. The designed phase shifter is constituted by an active single-to-differential convertor, a fully differential vector-summing core and a differential-to-single convertor. The mm-wave phase shifter, fabricated in 90 nm CMOS technology, provides 360° of phase shifting range over 35-43 GHz while consuming only 17.3 mW of power and 0.36 mm2 of chip area (excluding pads). Measurement results show that the phase shifter achieves an average gain of1.6-5 dB over 35-43 GHz. RMS phase error and RMS amplitude error for 32 measured phase states are less than 1 dB and 5.1°. The proposed phase shifter, which is able to cover the released frequency bands of n260, is a good candidate for 5G mm-wave applications.","PeriodicalId":166178,"journal":{"name":"2021 IEEE 4th International Conference on Electronics and Communication Engineering (ICECE)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127148064","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Dental X-ray imaging which shows the clear internal structure of teeth is essential for diagnosis. However, the image details in both dark and bright regions are often hard to distinguish due to the misoperation and the limitation of dental Xray imaging machines. In addition, the emergence of some realtime X-ray imaging machines has put forward higher requirements for the processing speed of enhancement systems. In this paper, an adaptive enhancement system consisting of a top control unit (TCU), an image quality feature extraction unit (FEU), a multilayer perception (MLP) unit and an adaptive image processing unit (IPU) is presented. Four image quality features as well as MLPs are proposed to detect image quality problems and control the image processing performance. For IPU, a novel fast contrast limited adaptive histogram equalization (FCLAHE) is proposed to accelerate the interpolation process. Modified guided filter, laplacian filter, gamma correction, and FCLAHE are integrated into the IPU. The proposed system is implemented in register transfer level (RTL) and demonstrated on field programmable gate array (FPGA). It runs at a clock frequency of 133 MHz and is capable of processing $1980times 1080$ images or videos with a high throughput of 127.35 Mpixels/s. Moreover, the proposed system offers the top image enhancement performance among the state-of-the-art implementations and its peak throughput is $37times$ higher than a personal computer (PC) with Core i7 8750H.
{"title":"An FPGA Based Adaptive Real-Time Enhancement System for Dental X-rays","authors":"Haoyang Sang, Junsong Zhang, Liyi Yao, Zhao Wang, Kunmeng Luo, Wumeng Yin, Meng You, Bo Gao","doi":"10.1109/ICECE54449.2021.9674312","DOIUrl":"https://doi.org/10.1109/ICECE54449.2021.9674312","url":null,"abstract":"Dental X-ray imaging which shows the clear internal structure of teeth is essential for diagnosis. However, the image details in both dark and bright regions are often hard to distinguish due to the misoperation and the limitation of dental Xray imaging machines. In addition, the emergence of some realtime X-ray imaging machines has put forward higher requirements for the processing speed of enhancement systems. In this paper, an adaptive enhancement system consisting of a top control unit (TCU), an image quality feature extraction unit (FEU), a multilayer perception (MLP) unit and an adaptive image processing unit (IPU) is presented. Four image quality features as well as MLPs are proposed to detect image quality problems and control the image processing performance. For IPU, a novel fast contrast limited adaptive histogram equalization (FCLAHE) is proposed to accelerate the interpolation process. Modified guided filter, laplacian filter, gamma correction, and FCLAHE are integrated into the IPU. The proposed system is implemented in register transfer level (RTL) and demonstrated on field programmable gate array (FPGA). It runs at a clock frequency of 133 MHz and is capable of processing $1980times 1080$ images or videos with a high throughput of 127.35 Mpixels/s. Moreover, the proposed system offers the top image enhancement performance among the state-of-the-art implementations and its peak throughput is $37times$ higher than a personal computer (PC) with Core i7 8750H.","PeriodicalId":166178,"journal":{"name":"2021 IEEE 4th International Conference on Electronics and Communication Engineering (ICECE)","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124840773","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-12-17DOI: 10.1109/ICECE54449.2021.9674392
Farouk Suleiman, Chris J. Hughes, E. Obio
In this paper we propose, a novel adaptative histogram matching method to remove low contrast, smeared ink, bleed-through and uneven illumination artefacts from scanned images of historical documents. The goal is to provide a better representation of document images and therefore improve readability and the source images for Optical Character Recognition (OCR). Unlike other methods that are designed for single artefacts, our proposed method enhances multiple (low-contrast, smeared-ink, bleed-through and uneven illumination). The method starts by taking the bimodal peaks of the original grayscale image and multiplying them to generated gaussian windows to create an ideal histogram with weights of importance to distribution. This histogram becomes the reference histogram to be matched to the original image for a more optimized image. Median filtering is also incorporated in the method to remove salt and pepper noise. We demonstrate the technique on the European Newspapers project (ENP) dataset chosen from the Pattern recognition and image analysis research lab (PRImA) and establish from the results that, the proposed method significantly reduces background noise and improves image quality on multiple artefacts as compared to other enhancement methods tested. To evaluate the efficiency of the proposed method, we make use of several performance criteria. These include Signal to Noise Ratio (SNR), Mean opinion score (MOS), and visual document image quality assessment (VDIQA) metric. The proposed method performs best in all the evaluation metrics having a 42.6 % increment on the average score of the other methods for MOS, 44.3% increment on average score of other methods for SNR and 61.11% better in VDIQA compared to other methods.
{"title":"Adaptive Enhancement for Scanned Historical Document Images","authors":"Farouk Suleiman, Chris J. Hughes, E. Obio","doi":"10.1109/ICECE54449.2021.9674392","DOIUrl":"https://doi.org/10.1109/ICECE54449.2021.9674392","url":null,"abstract":"In this paper we propose, a novel adaptative histogram matching method to remove low contrast, smeared ink, bleed-through and uneven illumination artefacts from scanned images of historical documents. The goal is to provide a better representation of document images and therefore improve readability and the source images for Optical Character Recognition (OCR). Unlike other methods that are designed for single artefacts, our proposed method enhances multiple (low-contrast, smeared-ink, bleed-through and uneven illumination). The method starts by taking the bimodal peaks of the original grayscale image and multiplying them to generated gaussian windows to create an ideal histogram with weights of importance to distribution. This histogram becomes the reference histogram to be matched to the original image for a more optimized image. Median filtering is also incorporated in the method to remove salt and pepper noise. We demonstrate the technique on the European Newspapers project (ENP) dataset chosen from the Pattern recognition and image analysis research lab (PRImA) and establish from the results that, the proposed method significantly reduces background noise and improves image quality on multiple artefacts as compared to other enhancement methods tested. To evaluate the efficiency of the proposed method, we make use of several performance criteria. These include Signal to Noise Ratio (SNR), Mean opinion score (MOS), and visual document image quality assessment (VDIQA) metric. The proposed method performs best in all the evaluation metrics having a 42.6 % increment on the average score of the other methods for MOS, 44.3% increment on average score of other methods for SNR and 61.11% better in VDIQA compared to other methods.","PeriodicalId":166178,"journal":{"name":"2021 IEEE 4th International Conference on Electronics and Communication Engineering (ICECE)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129545516","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-12-17DOI: 10.1109/ICECE54449.2021.9674434
Lu Dong, Yong Huang, Xi Chen
The fast development of the wireless technology has enabled phased-array technology and 5G communication technology a hot research point. As an indispensable module in the wireless technology, the radio frequency front-end chip influences the performance of the wireless system. Silicon-based integrated circuit has been attracting increasing attention in the micrometer and millimeter wave filed, because it has many advantages, such as low cost, low power consumption and high integration. In order to achieve amplitude control with large range and high precision, three silicon-based attenuator chips with different structures are proposed in this paper, and their simulation design and processing test are carried out. The test results are basically consistent with the simulation, and the performance of devices is excellent. Firstly, they can work in ultra-wide microwave frequency range $(mathrm{D}mathrm{C}sim 50mathrm{G}mathrm{H}mathrm{z})$. Secondly, the proposed attenuators feature very small size $(0.7mathrm{m}mathrm{m}^{star}0.7mathrm{m}mathrm{m}^{star}0.1mathrm{m}mathrm{m})$, which is conducive to the miniaturization of integrated circuits. These attenuators can be used in various circuits, whether in communication technology, radar phased control technology, radio frequency technology, or other electronic circuits, as long as there is an amplifier circuit, almost all of them can not do without attenuator.
{"title":"Novel Silicon-based Attenuator Chip","authors":"Lu Dong, Yong Huang, Xi Chen","doi":"10.1109/ICECE54449.2021.9674434","DOIUrl":"https://doi.org/10.1109/ICECE54449.2021.9674434","url":null,"abstract":"The fast development of the wireless technology has enabled phased-array technology and 5G communication technology a hot research point. As an indispensable module in the wireless technology, the radio frequency front-end chip influences the performance of the wireless system. Silicon-based integrated circuit has been attracting increasing attention in the micrometer and millimeter wave filed, because it has many advantages, such as low cost, low power consumption and high integration. In order to achieve amplitude control with large range and high precision, three silicon-based attenuator chips with different structures are proposed in this paper, and their simulation design and processing test are carried out. The test results are basically consistent with the simulation, and the performance of devices is excellent. Firstly, they can work in ultra-wide microwave frequency range $(mathrm{D}mathrm{C}sim 50mathrm{G}mathrm{H}mathrm{z})$. Secondly, the proposed attenuators feature very small size $(0.7mathrm{m}mathrm{m}^{star}0.7mathrm{m}mathrm{m}^{star}0.1mathrm{m}mathrm{m})$, which is conducive to the miniaturization of integrated circuits. These attenuators can be used in various circuits, whether in communication technology, radar phased control technology, radio frequency technology, or other electronic circuits, as long as there is an amplifier circuit, almost all of them can not do without attenuator.","PeriodicalId":166178,"journal":{"name":"2021 IEEE 4th International Conference on Electronics and Communication Engineering (ICECE)","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132459169","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In order to make full use of the concept of model-based systems engineering (MBSE) and the collaborative design capability of the open model-based engineering environment (OpenMBEE) in establishing complex product development platform, we propose an effective method to solve the problem of unified identity authentication in OpenMBEE integration. The proposed method is based on the Model View Controller(MVC) architecture in OpenMBEE, which has the characteristics strong plasticity and can be benefit for expanding the function in the controller layer. In our research, we make full use of the Software Development Kit (SDK) of OpenMBEE to expand the existing functions in the controller layer, and increase the function of communicating with the application development platform, then realize the function of sharing user authentication information between the development platform and OpenMBEE. After expansion, the system includes three modules, including the Front-end Service Module(FSM), MQ based Information Receiving Service Module(IFRSM), and the OpenMBEE Backend Server Module(BSM). The effectiveness of the proposed strategies is verified by some practical instances, which verifies that our study can provide an effective design idea for the identity authentication in OpenMBEE integration.
为了在构建复杂产品开发平台时充分利用基于模型的系统工程(MBSE)的概念和基于模型的开放工程环境(OpenMBEE)的协同设计能力，提出了一种解决OpenMBEE集成中统一身份认证问题的有效方法。该方法基于OpenMBEE中的模型-视图-控制器(Model - View - Controller, MVC)体系结构，具有可塑性强的特点，有利于控制器层功能的扩展。在我们的研究中，我们充分利用OpenMBEE的软件开发工具包(Software Development Kit, SDK)对控制器层已有的功能进行了扩展，增加了与应用开发平台的通信功能，实现了开发平台与OpenMBEE之间用户认证信息的共享功能。扩容后的系统包括FSM(前端业务模块)、IFRSM(基于MQ的信息接收服务模块)和BSM (OpenMBEE后端服务器模块)三个模块。通过实例验证了所提策略的有效性，为OpenMBEE集成中的身份认证提供了一种有效的设计思路。
{"title":"Key Techniques on Unified Identity Authentication in OpenMBEE Integration","authors":"Junjie Xue, Junhua Zhou, Guoqiang Shi, Chaoqun Feng, Penghua Liu, Hongyan Quan","doi":"10.1109/ICECE54449.2021.9674355","DOIUrl":"https://doi.org/10.1109/ICECE54449.2021.9674355","url":null,"abstract":"In order to make full use of the concept of model-based systems engineering (MBSE) and the collaborative design capability of the open model-based engineering environment (OpenMBEE) in establishing complex product development platform, we propose an effective method to solve the problem of unified identity authentication in OpenMBEE integration. The proposed method is based on the Model View Controller(MVC) architecture in OpenMBEE, which has the characteristics strong plasticity and can be benefit for expanding the function in the controller layer. In our research, we make full use of the Software Development Kit (SDK) of OpenMBEE to expand the existing functions in the controller layer, and increase the function of communicating with the application development platform, then realize the function of sharing user authentication information between the development platform and OpenMBEE. After expansion, the system includes three modules, including the Front-end Service Module(FSM), MQ based Information Receiving Service Module(IFRSM), and the OpenMBEE Backend Server Module(BSM). The effectiveness of the proposed strategies is verified by some practical instances, which verifies that our study can provide an effective design idea for the identity authentication in OpenMBEE integration.","PeriodicalId":166178,"journal":{"name":"2021 IEEE 4th International Conference on Electronics and Communication Engineering (ICECE)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114957207","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
More diverse applications and services pose a high demand for tracking services in indoor environments to improve user experience. Different from other positioning methods, the trajectory-based positioning system utilizes abundant historical information to further improve positioning accuracy. To better utilize historical information, we propose a novel historical information fusion method based on trajectory for indoor localization. Specifically, we first evaluate the distances between the reference points (RPs) and the previous position to match proper RPs. Then, a fusion weight is calculated according to the previous position and the change tendency of received signal strength. Based on the fusion weight, the position of target node can be determined. Finally, experiments are conducted and simulation results show that the positioning accuracy is improved significantly by the proposed algorithm.
{"title":"Improved KNN Algorithm with Historical Information Fusion for Indoor Positioning","authors":"Hui Zhang, Zhikun Wang, Yiyang Ni, Wenchao Xia, Haitao Zhao","doi":"10.1109/ICECE54449.2021.9674404","DOIUrl":"https://doi.org/10.1109/ICECE54449.2021.9674404","url":null,"abstract":"More diverse applications and services pose a high demand for tracking services in indoor environments to improve user experience. Different from other positioning methods, the trajectory-based positioning system utilizes abundant historical information to further improve positioning accuracy. To better utilize historical information, we propose a novel historical information fusion method based on trajectory for indoor localization. Specifically, we first evaluate the distances between the reference points (RPs) and the previous position to match proper RPs. Then, a fusion weight is calculated according to the previous position and the change tendency of received signal strength. Based on the fusion weight, the position of target node can be determined. Finally, experiments are conducted and simulation results show that the positioning accuracy is improved significantly by the proposed algorithm.","PeriodicalId":166178,"journal":{"name":"2021 IEEE 4th International Conference on Electronics and Communication Engineering (ICECE)","volume":"88 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133525598","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-12-17DOI: 10.1109/ICECE54449.2021.9674383
B. Doraswamy, K. Krishna, M. N. Giri Prasad
Drone technology is utilized for a variety of reasons, including military, agricultural, aerial photography, surveillance, remote sensing, and more. Based on real-time processing techniques, a drone plane is presented for monitoring and targeting public area crime theft in this proposed work. Previously crime prediction model was developed using Artificial Neural Network (ANN) and Regressive Neural Network (RNN), as they suffer from inappropriate accuracy levels and long-time computation. Thus, to overcome this drawback, Cat boost machine learning has been implemented as it uses tree-shaped primitives for the prediction that makes classification faster for the IoT environment. Buffalo-based Cat boosts Crime Prediction System (BCPS) initially collects crime data, preprocessing them, and then extracting environmental features and context features, the features are given to cat boost machine learning. The features are combined and give results as trees, and to improve accuracy, African Buffalo optimization (ABO) has been employed here. By estimating the predictors, a result has been obtained that was used for learning purposes and the testing side shows the result of crime theft detection. Thus BCPS is evaluated for results and compared with previous techniques to show the supremacy of the proposed model.
{"title":"Machine Learning Strategies for the Implementation of a Surveillance Drone","authors":"B. Doraswamy, K. Krishna, M. N. Giri Prasad","doi":"10.1109/ICECE54449.2021.9674383","DOIUrl":"https://doi.org/10.1109/ICECE54449.2021.9674383","url":null,"abstract":"Drone technology is utilized for a variety of reasons, including military, agricultural, aerial photography, surveillance, remote sensing, and more. Based on real-time processing techniques, a drone plane is presented for monitoring and targeting public area crime theft in this proposed work. Previously crime prediction model was developed using Artificial Neural Network (ANN) and Regressive Neural Network (RNN), as they suffer from inappropriate accuracy levels and long-time computation. Thus, to overcome this drawback, Cat boost machine learning has been implemented as it uses tree-shaped primitives for the prediction that makes classification faster for the IoT environment. Buffalo-based Cat boosts Crime Prediction System (BCPS) initially collects crime data, preprocessing them, and then extracting environmental features and context features, the features are given to cat boost machine learning. The features are combined and give results as trees, and to improve accuracy, African Buffalo optimization (ABO) has been employed here. By estimating the predictors, a result has been obtained that was used for learning purposes and the testing side shows the result of crime theft detection. Thus BCPS is evaluated for results and compared with previous techniques to show the supremacy of the proposed model.","PeriodicalId":166178,"journal":{"name":"2021 IEEE 4th International Conference on Electronics and Communication Engineering (ICECE)","volume":"46 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133479997","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-12-17DOI: 10.1109/ICECE54449.2021.9674628
Jie Yi, Jin Hou, Linxiao Huang, Haode Shi, Jian Hu
Although the present studies of face recognition have relatively been mature, in some complex scene environments, the efficiency of face recognition needs to be improved due to the influence of uncertain factors such as changes in illumination, changes in facial expressions, and partial facial occlusion. In order to improve the efficiency of face recognition, this paper proposes a feature fusion method based on convolutional neural networks (CNN) model and hog model. The model extracts rich implicit features from the original image by using convolutional neural network (CNN), and uses Dropout technology in the convolutional layer and the fully connected layer to randomly inhibit the activation of some neurons, so as to better solve the problem of overfitting. Moreover, this method also gives full play to the stability and robustness of Histogram of Oriented Gradients (HOG) Feature Enhancement Model. After extracting the CNN features and HOG features of the face, the method combines CNN SoftMax and HOG-SVM classifiers. The experimental results show that the recognition rate of this method is higher than that of single convolution neural network, which can reach 96.1%.
虽然目前人脸识别的研究已经相对成熟，但在一些复杂的场景环境中，由于光照变化、面部表情变化、部分面部遮挡等不确定因素的影响，人脸识别的效率还有待提高。为了提高人脸识别的效率，本文提出了一种基于卷积神经网络(CNN)模型和hog模型的特征融合方法。该模型利用卷积神经网络(CNN)从原始图像中提取丰富的隐式特征，并在卷积层和全连接层使用Dropout技术随机抑制部分神经元的激活，从而更好地解决过拟合问题。此外，该方法还充分发挥了HOG (Histogram of Oriented Gradients)特征增强模型的稳定性和鲁棒性。该方法在提取人脸的CNN特征和HOG特征后，结合CNN SoftMax和HOG- svm分类器。实验结果表明，该方法的识别率高于单一卷积神经网络的识别率，达到96.1%。
{"title":"Partial Occlusion Face Recognition Based on CNN and HOG Feature Fusion","authors":"Jie Yi, Jin Hou, Linxiao Huang, Haode Shi, Jian Hu","doi":"10.1109/ICECE54449.2021.9674628","DOIUrl":"https://doi.org/10.1109/ICECE54449.2021.9674628","url":null,"abstract":"Although the present studies of face recognition have relatively been mature, in some complex scene environments, the efficiency of face recognition needs to be improved due to the influence of uncertain factors such as changes in illumination, changes in facial expressions, and partial facial occlusion. In order to improve the efficiency of face recognition, this paper proposes a feature fusion method based on convolutional neural networks (CNN) model and hog model. The model extracts rich implicit features from the original image by using convolutional neural network (CNN), and uses Dropout technology in the convolutional layer and the fully connected layer to randomly inhibit the activation of some neurons, so as to better solve the problem of overfitting. Moreover, this method also gives full play to the stability and robustness of Histogram of Oriented Gradients (HOG) Feature Enhancement Model. After extracting the CNN features and HOG features of the face, the method combines CNN SoftMax and HOG-SVM classifiers. The experimental results show that the recognition rate of this method is higher than that of single convolution neural network, which can reach 96.1%.","PeriodicalId":166178,"journal":{"name":"2021 IEEE 4th International Conference on Electronics and Communication Engineering (ICECE)","volume":"32 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132929970","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：